|
| |
|
|
Noise reduction for low dose CT data
Holub, Zbyněk ; Odstrčilík, Jan (referee) ; Chmelík, Jiří (advisor)
The aim of this work is comparing of methods which are used for filtration of low dose CT images. This filtration is realized due to suppress of noise in final image and better visble of details. This work contains design and realization of selected filtration methods. Wiener´s correction coefficient and filtration with usage of WAVELET TRANSFORM are introduced in detail view. These methods are created in Matlab®; with the option to set para-meters of filteres. The filters are tested on 3D CT lungs image data. In the end of this work is a ranking of filter‘s quality and choosen of the most optimal approach.
|
|
|
Meta-analysis of bone tumorous lesions in spinal CT data using convolutional neural networks
Nantl, Ondřej ; Jakubíček, Roman (referee) ; Chmelík, Jiří (advisor)
This bachelor thesis deals with the use of convolutional neural networks in the meta-analysis of bone tumor lesions in CT image data. The theoretical part describes the anatomy and pathology of bone tissue, machine learning, discusses the functionality of convolutional neural networks and summarizes selected existing methods for computer-aided diagnosis of vertebra bone lesions. In the practical part, various types of models using convolutional neural networks were implemented and the networks were trained on an available augmented dataset. Finally, the results of various types of models were statistically evaluated, compared with available articles and discussed.
|
|
|
Detection of intracranial hemorrhages in head CT data
Nemček, Jakub ; Jan, Jiří (referee) ; Jakubíček, Roman (advisor)
This thesis deals with the detection of intracranial haemorrhages and their type classification in head CT images. The method of haemorrhages detection is based on a series of classifiers of the presence and type of haemorrhages in 2D CT slices in axial, sagittal and coronal plane, that may localise the bleedings and determine their types. The classifiers are based on the convolutional neural network architecture Inception-ResNet-v2. The head CT dataset CQ500 which is made available for public access, is used for the experiments. The thesis describes an additional manual annotation of the data, as the available annotations are insufficient for the purposes of the experiments. This thesis includes a theoretical basis of the essential medical knowledge, machine learning based classification and detection methods, and the detection algorithm proposal, realisation and testing. The algorithm performance is evaluated and discussed together with the potential implementation of the algorithm in computer-aided diagnosis systems.
|
|
|
Blood vessel tree segmentation of the mouse liver in CT data
Smékalová, Veronika ; Odstrčilík, Jan (referee) ; Kolář, Radim (advisor)
The methodology of visualization of soft tissue is in biology and medicine a topic for many years. During this period there were approving many techniques how to achieve accurate and authentic image of the researched object or structure. X-ray computed tomography is very helpful to get this goal but is necessary to improve contrasting techniques as well as the techniques of image post-processing. This thesis deals with imaging soft tissue. Specifically, it focuses on mouse liver contrasting with the artificial resin Microfil. Thesis also describes image processing technique (thresholding and region growing) for the data of the measurement with the goal of the visualization of the sample in 3D.
|
|
|
Segmentation of 3D medical images based on region growing method
Kantorová, Martina ; Krátká, Lucie (referee) ; Harabiš, Vratislav (advisor)
This bachalor thesis deals with a region growing approach for segmentation of volumetric medical images. The aim is to present basic methods of segmentation of image data and to focus in particular on the approach of region growing. The input data are brain slices of magnetic resonance imaging which can be visualized using the browser into the three basic planes. The viewer is implemented in MATLAB programming environment. Image segmentation is realized by seeded region growing.
|
|
|
Analysis of the use of 3D slicer SW for computational modeling in biomechanics
Kratochvílová, Hana ; Marcián, Petr (referee) ; Vosynek, Petr (advisor)
This thesis deals with a freeware program called 3D Slicer and its usefulness in the area of biomechanics. The introduction part describes the scope of the program and its functions. The next chapter shows a step-by-step modelling of a bone geometry using CT images and a comparison of geometries of bones created with different grayscale level using a program Gom Inspect. The last part focuses on importing geometries created in 3D Slicer, specifically femur and pelvis, into FEM platform ANSYS Workbench for preprocessing and defining femur’s stress and deformation characteristics.
|
|
|
Segmentation of bone lesions in spinal CT data
Zaťko, Martin ; Chmelík, Jiří (referee) ; Jakubíček, Roman (advisor)
The aim of the bachelor thesis was to get acquainted with the anatomy and oncological diseases of spine. Search for segmentation techniques and implement my chosen machine learning technique for the task of segmenting bone lesions of vertebral bodies. The U-net architecture of convolutional neural networks, which is generally widely used in the segmentation of biomedical images, was selected and implemented. The results obtained are high enough for the network to be used for initial rough detection and segmentation, but its use in the clinical world is not recommended.
|
|
|
Machine learning based method for medical image generation
Hrtoňová, Valentina ; Chmelík, Jiří (referee) ; Jakubíček, Roman (advisor)
This thesis deals with the use of generative adversarial networks for the synthesis of medical images. Firstly, artificial neural networks are described with a focus on convolutional neural networks and generative adversarial networks. Applications of generative adversarial networks in medicine are reviewed, and selected publications on the topic of medical image synthesis are described in more detail. Furthermore, multiple models of generative adversarial networks are designed and implemented in the Python programming language. First is a model of the deep convolutional generative adversarial network and the model „pix2pix“ for the generation of skin lesion images. Moreover, the „pix2pix“ model is used for the generation of both axial and sagittal CT images of the spine. Finally, the results of generating medical images using generative adversarial networks are presented and discussed.
|
|
|
Displaying 3D Graphics in Web Browser
Sychra, Tomáš ; Kršek, Přemysl (referee) ; Španěl, Michal (advisor)
This bachelor's thesis deals with display of accelerated 3D graphics in a web browser environment. Existing technologies such as WebGL are presented and discussed. Further, in the second part of the thesis, an application for browsing medical volumetric data is designed and implemented. The application is built with the WebGL technology and Javascript graphics engine called O3D API.
|
guest ::
